The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill requires the agent to execute npx -y @nomadamas/k-skill@0 instruct court-auction-notice-search as a primary step to obtain instructions. This mechanism allows the author to remotely define or modify the agent's behavior at runtime.
[REMOTE_CODE_EXECUTION]: The skill provides an update command npx -y @nomadamas/k-skill@0 update which executes remote code to maintain the skill and related components.
[EXTERNAL_DOWNLOADS]: The skill references and downloads instructions and helper files from an untrusted GitHub repository (github.com/NomaDamas/k-skill).
[INDIRECT_PROMPT_INJECTION]: The skill processes real estate auction data from courtauction.go.kr which is then ingested into the agent's context. This creates a potential attack surface where instructions embedded in the website content could influence agent behavior.
Ingestion points: Data is fetched via searchSaleNotices, getSaleNoticeDetail, and searchProperties (instruction.md).
Boundary markers: The skill includes instructions for the agent to warn users to verify data against the original court notices.
Capability inventory: The skill can execute shell commands via npx, perform network requests, and use Playwright for browser automation.
Sanitization: Input normalization is mentioned for case numbers, but there is no evidence of sanitization for the content returned from the external website.
[COMMAND_EXECUTION]: The skill documentation provides multiple CLI examples that involve executing the court-auction-notice-search command with various arguments to search for auction data.

court-auction-notice-search